Peptides of fewer than 100 amino acids that are directly translated from short open reading frames (sORFs) encoded in the genomes of widely divergent species, including human, have recently been discovered to have important biological functions, including prevention of disease-related apoptosis by the human peptide humanin. These genomically encoded peptides (GEPs) challenge not only the canonical proteolytic model of bioactive peptide production but also our understanding of what constitutes a gene. However, only a handful of human GEPs have been reported, so a complete catalog of these peptides is required before we can understand the full scope of their bioactivities. We will therefore apply a liquid chromatography-mass spectrometry (LC-MS)-based peptidomics approach to large-scale GEP discovery in human cell lines. Subsequent targeted genetic knock-downs and overexpression studies will be used to experimentally assign new GEPs to the short genes that encode them. We will also use the tools of chemical biology to selectively enrich GEPs from the cellular peptidome, permitting detection and discovery of low-abundance GEPs. Finally, our preliminary results reveal GEPs encoded in ORFs deep inside annotated RNAs, which are unlikely to be translated by ribosome scanning. We will therefore investigate the mechanism of translation initiation for these GEPs, and specifically determine if they are produced by internal initiation at internal ribosome entry sites. This project will not only provide fundamental new biological insights into mechanisms of bioactive peptide production and translation initiation, but will also provide a complete catalog of a new class of gene products with potential biomedical relevance.